Vertical Transportation System Power Usage: Behavioural Case Study of Regulated Buildings in Bangkok

Abstract

Sustainable urban development worldwide is crucial for the development of living spaces in high-rise buildings and infrastructures, which leads to the inevitability of increased energy consumption and demand of vertical transportation systems. The evaluation of the energy consumption of transportation systems is needed to verify and analyse the power usage related to traffic demands and patterns. In addition, efficient vertical transportation systems are central to the formulation of more sustainable cities. Therefore, this trend represents a substantial portion of the overall energy consumption of the building types. The benchmarking of the energy needs of the vertical transportation systems in five different building types via the comparison of granular load profile patterns (in conjunction with population densities) to the energy consumed was conducted, and it will be used to infer some impactful design strategies for the future. This study demonstrated a systematic approach to determine the power usage patterns in vertical transportation systems by actual measurement and traffic data collection from elevator monitoring. This may be used to develop a prediction for other cases in different types of installed vertical transportation systems. Therefore, the power usage of the vertical transportation systems can be used to determine the correlation between energy consumption and load pattern based on building characteristics and the overall energy consumption of each presented system.

1. Introduction

The electric elevator is an essential piece of equipment for transporting or moving people and goods to specific destinations in most of the world’s tall buildings. Not only are elevators utilized in skyscrapers to service the building, but they are also used in buildings with a small number of floors. Therefore, it is necessary to make use of the advantages of elevators; for example, as a necessity to assist people with disabilities. The energy use of public buildings accounts for a large proportion of the total building energy consumption. Building energy conservation is critical for Thailand’s energy conservation and reduction efforts. Therefore, considering energy conservation efforts in the construction sector, it is essential to perform energy audits of public buildings, evaluate their energy usage, and determine the key influencing elements. In the literature, several approaches using linear regression have recently been applied to hotel buildings for energy data benchmarking. However, the energy prediction was conducted based on a pre-defined methodology and up-to-date data, in order to reveal significant progress in the power generation sector [1]. Meanwhile, the residential sector’s energy consumption proportion across all sectors (i.e., commercial, residential, transportation, and industrial) increased from 14.2% in 1990 to 18.1% in 2007, as described in [2]. This indicates that apartment buildings, as a section of the private housing market, were becoming less energy efficient while the economy grew and the living standards improved; namely, people have come to prefer to consume more energy to have a more comfortable existence. However, behaviour changes in residents may not be an effective strategy for reducing energy consumption. Research work on the building sector in China found that it consumed 25% of the country’s total energy; in particular, public buildings consume more power than other types of buildings. For example, the energy consumption per square meter of non-residential buildings in China was three times that of residential buildings [3]. Other alternative methodologies, using the Lorenz curve and the Gini coefficient, have been implemented to investigate the energy consumption of office buildings in various cities in China [4]. This method serves as an alternative to the traditional procedure for differentiating human behaviour and building performance patterns. However, building energy efficiency has long been seen as a critical part of energy conservation. In-depth energy analysis for buildings can be conducted using various methods, such as the logarithmic mean Divisia index (LMDI) decomposition approach and the Hybrid LDMI for energy consumption and intensity. The additive decomposition approach is suitable when considering a quantity indicator [5,6]. Meanwhile, the passenger behaviour algorithm has been utilized for thermal simulation, the results of which were interesting and helped to justify the occupancy profile. However, there is still no relevant information regarding elevator systems [7]. The relationship between occupancy rate and energy usage intensity is difficult to identify. Understanding occupant behaviour creates new opportunities to influence the evolution of building technology, leading to increased energy efficiency and occupant comfort [8]. The high energy consumption and low energy efficiency pattern is prevalent in public buildings, implying that public buildings have significant energy-saving potential [9]. The function of a building type directly impacts the energy consumption per unit area [10]. Similar work has also been described in [11], comparing three types of public buildings. It was found that superstores consumed the most energy—more than double that of hotels and approximately five times that of government office buildings. Energy management and energy awareness programs promoting energy efficiency are expected to help increase hotel profits [12]. The use of the analytic hierarchy process and multi-perspective approach have been proposed for the evaluation and assessment of emerging renewable energy [13]. In the case of hotels, the energy consumption patterns of the building tend to differ from those of other building types [14]. The energy usage characteristics of hotel buildings have been studied, revealing that the electricity accounted for 75% of the total energy consumption. In the case of hospitals and schools, the average values of energy consumption and carbon emissions per unit area for different building types, such as offices, hospitals, and schools, have been determined [15]. Therefore, a vast number of studies have been conducted in order to reduce the gap between the predicted and actual energy consumption in buildings; however, there remains some doubt [16]. From the literature, it is apparent that the energy consumption of hospitals is significantly higher than that of government offices and schools. However, these impact assessments often do not consider vertical transportation and the associated apparatus. Measuring and monitoring energy consumption in buildings becomes essential to understanding the actual energy consumption, thus providing designers and building operators feedback to promote building energy efficiency [17]. This technique, however, requires significant time for data collection. Through measurement and calculation, the energy consumption of elevators based on the hybrid approach has been presented for low- and high-rise buildings [18,19]. This method collects information from the management office based on daily energy use, from which the elevator’s energy efficiency classification can be established. The results indicate that the low-rise building had better energy efficiency than the high-rise building. Nevertheless, the classification method was modified to VDI4707 in order to measure the elevator’s energy efficiency, which can provide a suggestion to improve the energy consumption of the building, as described in [20]. Several methods for evaluating the comprehensive operations of elevators have been reported in the literature [21,22]. These evaluations are related to data collection through measurement and data analysis. The energy consumption of a conventional drive in an elevator has been compared to that of a regenerative drive, as described in [23]. Differences between building types and elevator performance assessments lead to distinct performance levels, related to many variables, measurements, and performance evaluations [24,25]. This can possibly be thought of as an alternative route to achieve a potential reduction of the standby energy consumption. In general, standby mode represents 5–95% of the energy consumption. The standby energy can be reduced by 80% if drive technology is implemented [26]. Building energy saving through elevators by adaptive counterweight sizing has been proposed, involving the measurement of daily elevator traffic and load behaviour, through which significant savings can be achieved [27]. Due to the problems of the traction wheel and the wire rope slippage causing excessive wear and tear, the image processing technique has been adapted to detect wire rope slippage, which can enhance safety usage and reduce energy consumption [28]. To establish the energy efficiency rating of an elevator installation with sufficient confidence, the energy performance and operating parameters must be measured under normal service conditions [29]. Interestingly, the occupancy behaviour is inherently complex and stochastic, making it difficult to accurately describe or predict building energy performance [30]. Therefore, this paper’s main aim is to analyse traffic behaviour and evaluate the energy consumption of five regulated building types in Thailand: namely office, university, hospital, hotel, and residential.

In summary, the key contributions of this paper are as follows:

This study analyses the elevator load pattern and population density for different types of buildings and determines the correlation between energy consumption and load pattern, based on building characteristics.

The remainder of this paper is organized as follows: Section 2 provides the principle of vertical transport power usage behaviour. The methodology for behavioural analysis of the vertical transport power usage is presented in Section 3. The results and discussion of the aforementioned behavioural analysis are presented in Section 4. Finally, Section 5 presents the conclusions.

2. Principle of Vertical Transport Power Usage Behaviour

In a vertical transportation system, symmetry refers to the reflection of a pattern at an angle across a centreline, called an axis. Commercial buildings are used for commercial purposes, and are comprised of various business types, office buildings, stores, industrial buildings, and self-parking garages. Apartments and hotels are considered residential buildings, while schools and hospitals are considered institutional buildings. Commercial buildings have definite transportation requirements, as the arrival and departure of their populations are normally focused within certain periods of the working day. Depending on the operating direction and several other conditions, an elevator can operate in four quadrants, as shown in Figure 1. When the combined weight of the car and its load is less than the counterweight, the load is considered to be light condition. The operating mode of an elevator is based on the load conditions and direction, which has defined by individual quadrant Q1, Q2, Q3 and Q4 represented in Figure 1, (Q4) travelling upward with a heavy load and downward with a light load (Q1) results in the motoring mode; that is, instances in which the motor consumes energy from the power source.

On the other hand, travelling downward with a heavy load (Q3) and ascending with a light load (Q2) results in the motor entering generating mode, thus generating power. In a conventional elevator, the generated power is dissipated as heat by the breaking resistor, resulting in energy loss. It is considered undesirable to conserve this portion of energy and reuse it in subsequent operating cycles, even though this would result in significant energy savings [23]. As represented in Figure 2, the power meter is directly connected to the power supply, which directly connects to the rectifier unit, connecting the DC-link circuit through to the inverter unit and supplying power to the motor. The motor’s shaft is directly connected to the driving sheave and the diverter sheave. The driving sheave has a steel wire rope suspended on it, with one end attached to the car and the other end connected to the counterbalance.

This section describes our model of calculating the total energy consumption, which is applicable to direct calculation, simulation, or a monitoring system. The total energy consumption is the sum of the running energy, ${\mathit{E}}_{\mathit{R}}$, and the standby energy, ${\mathit{E}}_{\mathit{S}}$:

$${\mathit{E}}_{\mathit{sd}}{=\mathit{E}}_{\mathit{R}}{+\mathit{E}}_{\mathit{S}}$$

(1)

where $E$ is the total energy consumption. More detail is provided in Equation (2). The integration of the instantaneous line power, ${\mathit{P}}_{\mathit{line}}$, provides the energy consumed within $t$, seconds:

$$\mathrm{w}={{\displaystyle \int }}_0^{\mathrm{t}}{\mathit{P}}_{\mathit{line}}\left({\mathit{P}}_{\mathit{m}}{+\mathit{P}}_{\mathit{l}}{+\mathit{P}}_{\mathit{c}}{+\mathit{P}}_{\mathit{b}}{+\mathit{P}}_{\mathit{i}}\right)\mathit{dt}$$

(2)

where ${\mathit{P}}_{\mathit{m}}$ represents the motor power; ${\mathit{P}}_{\mathit{l}}$ denotes the power losses of the motor; ${\mathit{P}}_{\mathit{c}}$ represents the control system and lighting which consume constant power; ${\mathit{P}}_{\mathit{b}}$ is the constant power of the brake; and ${\mathit{P}}_{\mathit{i}}$ denotes the invertor losses [31]. Generally, the components of losses are positive. Furthermore, the motor power can be positive or negative, depending on the direction of the energy flow. As a result, the line power may be harmful as well, revealing that the system is regenerating.

$${\mathit{E}}_{\mathit{sd}}=\left(24-\left(\frac{{\mathit{n}}_{\mathit{d}}}{3600}\right){+\mathit{t}}_{\mathit{av}}\right)\left({\mathit{P}}_{\mathit{id}}{\mathit{R}}_{\mathit{id}}{+\mathit{P}}_{\mathit{st}}{\mathit{R}}_{\mathit{st}}\right)$$

(3)

where ${\mathit{E}}_{\mathit{sd}}$ is the daily standby energy consumption, consisting of the two main components representing when energy is consumed and when the elevator is not in use; ${\mathit{R}}_{\mathit{id}}$ is the idle time ratio (when consuming, ${\mathit{P}}_{\mathit{id}}$ < 1); ${\mathit{R}}_{\mathit{st}}$ is the standby time ratio (when consuming, ${\mathit{R}}_{\mathit{st}}$ < 1); and ${\mathit{t}}_{\mathit{av}}$ represents time to travel, including door time, for each of the building types. An elevator that is not running can be expressed by Equation (4):

$${\mathit{t}}_{\mathit{av}}=\frac{{\mathit{S}}_{\mathit{av}}}{\mathit{v}}+\frac{\mathit{v}}{\mathit{a}}+\frac{\mathit{a}}{\mathit{j}}{+\mathit{t}}_{\mathit{d}}$$

(4)

where ${\mathit{t}}_{\mathit{av}}$ is the time to travel, including door time, for each of the building types. ${\mathit{S}}_{\mathit{av}}$ is the average distance. $v$ is the rated speed ($\mathrm{m}/\mathrm{s}$) and $\mathit{j}$ is the rated jerk (${\mathrm{m}/\mathrm{s}}^2$) while ${\mathit{t}}_{\mathit{d}}$ is the running time per day (hours). When operating the motor, the input power needs to be slightly more than the output, due to motor losses:

$${\mathit{P}}_{\mathit{line}}=\frac{{\mathit{P}}_{\mathit{out}}}{\eta }$$

(5)

Equation (6) represents that regenerating the motor will return slightly less energy than that generated, due to the motor loss:

$${\mathit{P}}_{\mathit{line}}{=\mathit{P}}_{\mathit{regen}} \times \eta$$

(6)

The total efficiency is presented in Equation (7), where $\eta$ represents the efficiency of the shaft, ${\eta }_{\mathit{shaft}}$, multiplied by the efficiency of the motor, ${\eta }_{\mathit{motor}}$:

$${\eta =\eta }_{\mathit{shaft}}{\times \eta }_{\mathit{motor}}$$

(7)

The typical power curve of the hoisting machine, as a function of the full load/full speed up power ${\mathit{P}}_{\mathit{norm}}$, can be presented as Equation (8), where $\mathit{Q}$ is the rated load (kg), $\mathit{V}$ is the rated speed (m/s), and ${\mathit{G}}_{\mathit{n}}$ is the acceleration due to gravity (equal to 9.81 ${\mathrm{m}/\mathrm{s}}^2$):

$${\mathit{P}}_{\mathit{norm}}=\frac{{\mathit{Q}\times \mathit{V}\times \mathit{G}}_{\mathit{n}}}{{2\times \eta \times \eta }_{\mathit{motor}}\times 1000}$$

(8)

During heavy traffic, the carload and the number of elevators starts are increased, while the travel distances decrease:

$$P_{Flow}=\left\{\begin{array}{c}Carload\le CWT; \\ Carload\ge CWT; \\ Carload\ge CWT; \\ Carload\le CWT; \end{array}\begin{array}{l}Elevetor\in Generator\wedge Upcontrol\\ Elevetor\in Generator\wedge Downcontrol\\ Elevetor\in Motor\wedge Upcontrol\\ Elevetor\in Motor\wedge Downcontrol \end{array}\right.$$

(9)

Figure 3 represents the motor’s operation from the initial state until constant speed of the elevator is achieved, travelling upward to the destination position that a passenger has assigned. In this initial state, the frequency is increasing. Until the slowdown to stop at the landing, the power consumption in each running period varies with the travelling distance. However, in this process, different motoring values occur, denoted by W1, W2, and W4 in the figure. W5 is not displayed in the diagram, as it describes a condition in which the motor is not active and does not consume power (or only consumes a minimal amount).

At the same time, the motor consumes electrical power. This working condition is called monitoring. The travelling time is shown in the t1 to t4 periods, respectively. tc represents the time from which the elevator starts running until it stops at the destination.

On the other hand, suppose the carload is greater than the counterbalanced weight when the elevator is travelling downward. In this state, the car moves downward, and the elevator does not use power from the electrical system. The motor is now operating in generator mode, with different values (as indicated by W6–W9). These vary by distance and appear over the period from t6 to t9, until the elevator stops. This condition is called generating mode.

3. Methodology for Behavioural Analysis of the Vertical Transport Power Usage

Onsite measurement methods can be adapted from the standard EN ISO 25745-1:2012 [21], which focuses on electricity consumption measurement procedures. The elevator usage is categorized according to the VDI 4707 standard which has provided five usage categories according to the usage pattern of an investigated elevator, the average travel and idle times, and the investigated object’s actual location.

Figure 4 shows the main distribution panel found in a typical setup, as seen in Figure 4a, and the power meter inside the elevator controller cabinet when some space is available for place power analyser, as seen in Figure 4b.

Figure 5a shows a photographic image of a connection point in a distribution panel located in an elevator machine room. This connection is mainly intended for the single-lift unit measurement for the proposed comparison. Inside the distribution panel, the space is limited; therefore, the power analyser as shown in Figure 5b needs to be located outside the distribution cabinet.

Figure 6a shows a photographic image of a geared traction elevator system, powered by induction motors and using a 1:1 roping ratio, installed in one of our sample hospital buildings. Figure 6b represents the hospital resistor used for braking. When the elevator operates in generator mode, the reverse excess energy is eliminated by the braking resistor in the form of heat, resulting in energy loss. In effect, this causes the temperature inside the elevator machine room to have a higher value, which is commonly the case for old buildings (e.g., aged at least twenty years).

Figure 7 shows the elevator network diagram and the monitoring system. Traffic data and elevator trips information can be extracted from the elevator monitoring system in a spreadsheet format.

Figure 8 represents the workflow diagram of the study process. After a conceptual study, the sampler site needs to be selected to conform with the building type. This system incorporates two important techniques: first, the electrical instrumental measuring technique and second, the curve fitting technique, which will provide most intriguing findings and equations for further analysis.

4. Results and Discussion

This section aims to outline the actual measurement results and the elevator information data. The resources of measurement data were taken for each building type from the elevator monitoring system and the traffic monitoring tool, as shown in Figure 7. The building database was extracted to numbers formatted and categorized to the following types: incoming, outgoing, and inter-floor. Traveling trip information, including directional and passenger loads, were arranged for curve fitting and further analysis. In this study, the occupancy rates were assumed based on the basic building design. The energy consumption is relatively proportional to the number of people using the elevators in each type of building. In accordance with ISO 25745-2:2015, the elevator usage was categorized based on the number of elevator trips. Therefore, the main achievements, including the contributions, are summarized as shown in

Table 1. Population density and building characteristics.

Type	Office	University	Hotel	Hospital	Residential
Usage category	6	5	5	6	4
No. of trips per day (trip)	2000	1500	1000	2500	900
GFA (m2)	80,000	50,000	28,000	70,262	40,050
Population Density	10 m2/person	18 m2/person	23 m2/person	21 m2/person	45 m2/room
Occupancy Rate	85%	95%	64%	83%	75%

.

Table 1 shows the number of trips per day for each type of building as a building characteristic. Meanwhile, for the buildings where elevator monitoring is not available, the elevator usage reports need to be download by the proprietary tool from each elevator company. Important components of the study are data collection and analysis together with the data integration technique. However, the impact of population density and building characteristics are related to data variation in the actual measurement system.

Figure 9 shows an office traffic pattern based on the results, and we make the following observations. Data for the office are denoted by solid yellow, purple, and green areas in the figure. The presence of two intense peaks in the plots corresponds to typical office passenger behaviour: the first peak (at approximately 08:00) represents incoming passengers, while the following peaks in the plot correspond to the lunch (at approximately 12:00 and 13:00) and evening (at approximately 17:00) peaks. Typically, an office building can be defined as single-tenant or multi-tenant, which will have an impact on the internal traffic of the building. Regarding passenger traffic in the building, we defined traffic flows as follows: outgoing, inter-floor, and incoming. Incoming traffic usually comprises 75–80% of the total population, while outgoing traffic comprises approximately 20%; the remaining proportion will be inter-floor traffic. In the case of offices, as single-tenant building will have more inter-floor traffic, due to the co-ordination between departments or the travelling between floors. During the lunch peak, the number of passengers increases, as they go out for lunch in the same direction, and in the evening, passengers begin to use elevators to exit the building, thus increasing the frequency from 17:00 to 18:00, which is typical behaviour for an office building.

Figure 10 shows that the university buildings are classified as institutional buildings. For example, a university campus may consist of office buildings, residence halls, catering facilities, and factory-like units for housing, teaching, and research equipment. In addition, certain buildings serve multiple functions, including lecture halls, laboratories, and offices. In particular, when universities are located in city centres, many have tall buildings (10–20 stories). Before and after each 50-min lecture, tutorial, or seminar session, there may be hourly cycles of 10 min of demand. These peaks can range between 15% and 25% with a 30–50 s interval. It is improbable that an economically feasible solution may be found to meet such high peak requirements, although extensive stair use is probable. The activity levels are low between the peaks. Office-type structures can be treated the same way as described above. In addition, low-rise research buildings may be required to comply with special movement requirements associated with the system components. Elevators are critical for guest and service staff circulation within a hotel. Escalators should be used for short-distance transfers, such as connecting function levels to the lobby. The average occupancy rate of a hotel varies according to its type. For example, it may be reasonable to assume a ratio of 1 person in business hotels, 1.5 people in transit hotels, and 2 people in vacation hotels. Hotel traffic patterns are complicated, compared to the morning and afternoon peaks in an office building, wherein check-out (08:00–11:00) and check-in (15:00–19:00) are busy periods. The high volume of two-way traffic is explained by guests travelling to and from their rooms and restaurants, among other places, either inside or outside the hotel. The arrival rate is dependent on the hotel’s star rating. We assume a 10% (1-star) to 15% (5-star) arrival rate and a 30 to 50 s interval (5- and 1-star, respectively). Calculations are made based on the assumption that there are an equal number of upward and downward stops in these cases. At most times, it is unlikely that the elevator car will be loaded at more than 50% of its capacity. However, hotel elevators should have a capacity of at least 1600 kg, in order to accommodate luggage and provide guests with an uncrowded and comfortable travel experience. This rule should be applied cautiously, as it may not apply to a low-rise hotel with 30% of its rooms on the ground level.

Figure 11 shows the relative arrival rates for hotels. There are two traffic peaks in the evening, when people check in and have dinner. The widths and the heights of the peaks and the proportions of incoming, inter-floor, and outgoing components may vary according to the building layout and the culture. The inter-floor traffic is between typical floors, including gyms, restaurants, and business centres. In the measured hotels, the maximum arrival rate was 9.5% of the population in five minutes; however, this may be inappropriate for a small-footprint high-rise hotel. Additionally, there are distinct operational requirements between transit hotels near airports or other locations where guests tend to stay for a single night as well as hotels used by longer-term and vacation guests. The structure’s layout is crucial; for example, whether it has a small footprint but is tall, or a vast footprint but is low. The vertical transportation system uses elevators as the main element for transportation in the building. Their effective operation is even more critical when dealing with hospitality situations. Although many city hospitals feature high-rise aspects, most hospitals in the provinces are developed on a low-rise premise, being less than 21 m in height. In addition, elevators in low-rise hospitals are primarily supplied to assist staff and visitors in moving bed-bound patients and performing service activities from floor to floor.

Figure 12 shows the hospital traffic pattern. When considering hospitals, the following factors are important: the number of staff and their shift patterns; the number of visitors and their hours of visitation; the location of theatres, X-ray departments, and other facilities; the distribution and delivery of food, beverages, and housekeeping supplies; waste disposal; patient emergency evacuation; and porters. To enhance infection control, it may be required to segregate patient bed elevators from visitor and staff elevators. The demand can be estimated by multiplying the number of beds by three, in order to account for staff, guests, and other factors. For example, in the case of hospital-type buildings, a reasonably large volume of travel between floors can be observed (as represented by the purple area), which is greater than the population leaving the building due to the transport of people between the floors and the contact between departments, similar to the case of a single-tenant office building. The incoming passengers are patients, visitors, and nurses travelling between departments. We can observe considerable variability throughout the period from the start of the service around 08:00 until 2:00 p.m., which is the time at which patient visits cease. A reasonable arrival rate is 8–10% with a 30–50 s delay; these numbers can be used to forecast future demand.

Figure 13 shows the residential pattern. The elevators can have rated loads ranging from 630 to 1000 kg. With a normal journey time of 30 s, the rated speed may be lower than that in hotels. The assessment of the residential population is based on the number of bedrooms and is heavily influenced by culture. For the first bedroom, Barney has recommended 1.5–2 people and, for subsequent bedrooms, 0–5 people [32]. When the configuration and use of the residential floor are unknown, Strakosch has recommended either 1.5–2 persons per bedroom or 20 m² net area per person [33]. Both residential and hotel building types are mostly characterized by two-way traffic, but hotel traffic is more intense. The highest traffic peak for the residential building was 5.7% in five minutes. In residential buildings, there is a morning downtime, and night-time incoming traffic is slightly more common, while daytime inter-floor traffic is negligible. The individual building measurements were combined into average and worst-case profiles for planning purposes. The average of all arrival rates was determined for the average profile and, for the worst-case profile, the highest arrival rate among the four measurements for each period was chosen.

Figure 14a shows the relationship between the elevator’s travel distance versus the carload and energy consumption in the office building, indicating a significant amount of energy consumption. With increasing carloads and travel distances, two intense peaks appear in the plots, corresponding to higher power consumption. Figure 14b represents the relationship of carload versus increasing energy consumption, which indicates the energy consumption pattern per carload as a polynomial equation with a variance of 0.89443, as presented in Equations (10) and (11). From Equation (10), it can be determined that, while the elevator is in standby mode and there are no moving parts, the energy used for the control system has a value of 0.00116.

$${\mathrm{E}}_{\mathrm{p}}=-{2.90407 \times 10}^{-7}{\mathrm{L}}^2 + {5.03707 \times 10}^{-4} \mathrm{L} + 0.00116$$

(10)

$${\mathrm{R}}^2=0.89443$$

(11)

The demand for elevators to access parking lots is primarily determined by the maximum rate of vehicle entry and exit and the average occupancy of each vehicle. These values can be determined through a traffic study.

Figure 15a shows the relationship between the elevator travel distance versus the carload and the energy consumption in a university building, presenting a significant amount of energy consumption. As mentioned above, additional carloads and increasing travel distances led to two intense peaks in the plots, indicating higher power consumption. Figure 15b represents the relationship between carload and energy consumption, which indicates the energy consumption pattern per carload as a polynomial equation, with a variance equal to 0.89674, as presented in Equations (12) and (13). When the elevator is in standby mode and there are no moving parts, the energy used for the control system has a value of 0.00655.

$${\mathrm{E}}_{\mathrm{p}}=-{2.72985 \times 10}^{-7}{\mathrm{L}}^2+{5.15896 \times 10}^{-4}\mathrm{L}+{6.55258\times 10}^{-4}$$

(12)

$${\mathrm{R}}^2=0.89674$$

(13)

Figure 16a presents the relationship between the elevator’s travel distance versus the carload and energy consumption in the hospital building, indicating a significant amount of energy consumption with higher carloads and travel distances. Figure 16b represents the relationship between carload and energy consumption, which indicates the energy consumption pattern per carload as a polynomial equation with a variance equal to 0.92964, as presented in Equations (14) and (15). When the elevator is in standby mode and there are no moving parts, the energy used for the control system has a value of 0.00247.

$${\mathrm{E}}_{\mathrm{p}}=-{1.12623 \times 10}^{-5}{\mathrm{L}}^2+0.00142\mathrm{L}+{2.4780\times 10}^{-4}$$

(14)

$${\mathrm{R}}^2=0.92964$$

(15)

Figure 17a presents the relationship between the elevator’s travel distance, carload, and the energy consumption in the hotel building, indicating a significant amount of energy consumption. With increasing carloads and travel distances, two intense peaks can be observed in the plot, corresponding to high power consumption. Figure 17b represents the relationship between carload and energy consumption, which indicates the energy consumption pattern per carload as a polynomial equation with a variance of 0.89901, as presented in Equations (16) and (17). When the elevator is in standby mode and there are no moving parts, the energy used for the control system has a value of 0.77985.

$${\mathrm{E}}_{\mathrm{p}}={3934.360139\mathrm{L}}^2+1581.92603\mathrm{L}-0.77985$$

(16)

$${\mathrm{R}}^2=0.89901$$

(17)

At present, Bangkok has limited space for residential project development. Therefore, most residential project development in Bangkok is high-rise development; namely, condominium projects. Condominiums can respond to the daily living habits of the people in Bangkok very well in terms of convenience of traveling. Moreover, the number of facilities in these projects, such as vertical transport systems, has correspondingly increased.

Figure 18a shows the relationship between an elevator’s moving distance, carload, and energy consumption in a residential building, indicating again a significant amount of energy consumption. With increasing carloads and travel distances, two intense peaks in the plots appear, corresponding to high power consumption. Figure 18b represents the relationship between carload and increasing energy consumption, which indicates the energy consumption pattern per carload as a polynomial equation with a variance equal to 0.93577, as presented in Equations (18) and (19). When the elevator is in standby mode and there are no moving parts, the energy used for the control system has a value of 0.000733.

$${\mathrm{E}}_{\mathrm{p}}=-{6.30654 \times 10}^{-7}{\mathrm{L}}^2+{4.57821 \times 10}^{-4}\mathrm{L}+{7.33511\times 10}^{-4}$$

(18)

$${\mathrm{R}}^2=0.93577$$

(19)

Table 2. Elevator characteristics with respect to various building types.

Type	Office	University	Hotel	Hospital	Residential
Rated load (kg)	2000	1600	1800	1800	1800
Rise (m)	75	68	87	65	200
Number of floors	24	19	24	12	50
Nominal speed (m/s)	2.5	2.0	2.5	2.0	6.0
Acceleration (m/s2)	1.0	1.0	1.0	0.8	1.0
Jerk (m/s3)	1.2	1.2	1.2	0.8	1.2
Roping	2:1	2:1	1:1	1:1	1:1
Daily trips	2000	1500	1000	2500	900
Average carload (%)	40%	50%	30%	50%	25%
Counterbalancing (%)	50.00%	50.00%	50.00%	50.00%	50.00%

provides the specifications for each of the elevators considered in this study. These data reveal the variation in energy consumption between each different building type. The activity of the elevators has an impact on the trip pattern and directly affects the energy consumption behaviour. Therefore, the combined actual measurement data and elevator traffic information analysis is useful for estimating the energy consumption based on elevator weight loads; however, the data correction and analysis did not cover stair usage, due to the unpredictable behaviours of users and the limitations of the measurement tools. In the morning, an office building elevator will carry almost all passengers up and empty the car, and then come down to collect the passengers who are travelling in both upward and downward directions. In this period, the elevator will use power from the power supply to operate. At noon, the elevator will carry almost all passengers down, then run an empty car to pick up more passengers. As the passengers travel primarily in a downward direction, the elevator will return power to the electrical system. The elevator is fully loaded to pick up passengers in the afternoon, running the empty car to pick up more passengers in both the upward and downward directions. In this period, the elevator must be powered by the electrical system to operate. The elevator loads are almost full of passengers in the evening, which are run down while an empty car returns upwards. Therefore, in both directions, the elevator returns electricity to the supply. Analyses of elevator load patterns and the population densities related to different types of buildings are represented in Equations (10), (12), (14), (16) and (18), while the associated relationships between energy consumption and load pattern by building type are presented in Equations (11), (13), (15), (17) and (19). Moreover, at any time, the elevator is used between floors. The electrical power consumption of the elevator in this case depends on the payload and the distance travelled over time, which determines the energy cost associated with electricity consumption. Alternatively, elevator systems with a smaller capacity may start up less frequently, while typically handling a higher relative load (i.e., of passengers). This means that the savings obtained through [14] counterweight resizing and utilizing a regenerating effect may differ slightly from those determined in this study. In the case of heavy traffic, carloads increase, which in turn increases the number of elevators starts and decreases the travel distances, as represented by Equation (9). Therefore, the polynomial equation derived for each of the building types can be used to predict, in terms of various loads, the energy consumption of each elevator from each building type. Interestingly, the polynomial equation from the curve fitting of the collected data can be further implemented to identify the weights of the passengers and to help identify the energy usage of the elevator.

5. Conclusions

In conclusion, this study highlights the challenge of the vertical transportation power usage assessed by actual measurement and data collection from the monitoring system. This work adapted the curve fitting technique to analyse the energy consumption of the vertical transportation system from each building. The building characteristics consist of the elevator load pattern and the population density for different types of buildings for which the energy consumption was revealed in this study. Moreover, the energy usage of the vertical transportation system differs on the account of various structures having differing numbers of stories and travel distances. Finally, this study demonstrated a systematic approach for determining the power usage patterns in vertical transportation systems by measurement data. The measurement data analysis can represent the power consumption by using the curve fitting technique. The polynomial equation from the curve fitting of each building type may be used as a predictor for the power consumption in another building type. Future research in conjunction with non-intrusive load monitoring (NILM) in the elevator system will be implemented in the near future.